Quote Check Agent · Use case
How can we use an AI agent to check a draft quote against our price book and discount policy on our own infrastructure?
A quote check agent takes one draft quote, reads your price book and your discount policy, and sets out, line by line, the list price, the price quoted, the discount, the rule of the policy that applies, and whether the line needs an approval and from whom. The approver named in the policy approves or refuses in your own system. The agent sets no price, proposes no discount, recommends neither approval nor refusal, changes nothing on the quote and issues nothing to the buyer.
The quote has forty lines.
The discount policy has its own.
- ✗ 01A salesperson with a draft quote reads the price book and the discount policy to find out which lines need an approval, and from whom, before the quote can go anywhere.
- ✗ 02A price that reaches a buyer is one the company is held to, so the check and the approval are not steps to hand to software.
Where it stops for a person
Always. The approver named in the policy approves or refuses.Always. It checks, it does not decide: it does not say "approve" or "refuse", and it suggests no other price. Where the policy leaves room for judgement it says so and quotes the clause. A product that is not in the price book, and a term the policy does not cover, are listed for the deal desk. The check is of the quote, not of the person who drafted it: it keeps no tally of who asks for which discount.
It lists what needs an approval.
The approver named in the policy decides.
Step · 01
A salesperson or the deal desk names the draft quote.
Step · 02
The agent reads the quote through your own APIs, read only: products, quantities, prices, discounts and the term.
Step · 03
It finds each product in the price book and the rules of the discount policy that apply, and quotes them.
Step · 04
It sets out each line: list price, quoted price, discount, the rule, and whether an approval is needed and whose. It lists what it could not find in the price book.
Step · 05
The approver named in the policy approves or refuses in your own system, and the salesperson issues the quote.
Reads from · connectors
Your own APIs
The one draft quote named, read through your own APIs with the operations that change something left out; and the price book, the discount policy and the approval matrix, indexed as a knowledge source. It runs with the permissions of the person using it.
Acts through · MCP servers
Nothing. It has no tool that changes a quote, sets a price or a discount, starts an approval, or sends a quote to anyone.
What you set up
- Index the price book, the discount policy and the approval matrix as a knowledge source, and remove the versions no longer in force.
- Import the description of the API that holds quotes, and keep only the operations that read.
- Build the agent in Katonic Studio and test it on past quotes where you know which approvals were needed.
- Publish it to one sales team and the deal desk first.
Where it runs
The agent, its knowledge store and the model run on infrastructure you own or control. You choose the model and serve it yourself, so questions, documents and answers are not sent to a public AI service. Quotes, prices and the discount policy stay inside the company, and every approval is given by a person in your own system.
How you get it today
This page is labelled Use case: the agent is built in Katonic Studio. You describe the job, connect what is named above, add test questions and release the agent once it passes them. When the Catalog holds a tested template for this job, this page will carry the Template label and show how to take it.
Does it set or change a price?
No. It reads the quote and has no tool that changes it. It sets no price and proposes no discount.
Does it approve a discount?
No. It lists each line that needs an approval and names the approver the policy gives. That person approves or refuses.
What if a product is not in the price book?
It says so and lists the line for the deal desk. It does not work out a price of its own.